Metformin for Weight Loss in PCOS: The Real Average, Not the Anecdote

How much weight does metformin actually cause you to lose?

A 2023 meta-analysis that informed the current international PCOS guideline pooled the randomised trials and found metformin produced a BMI reduction of 0.53 kg/m² (95% CI −0.95 to −0.12) compared with placebo. On a 160 cm frame that converts to roughly 1.4 kg; on a 175 cm frame, roughly 1.6 kg. That is the number the current evidence actually supports — not the double-digit transformation the drug is sometimes discussed as online.

This is a pooled average across dozens of trials, most of which combined metformin with lifestyle change rather than testing the drug alone. Individual results inside that average ranged from no change to several kilograms — the confidence interval on the pooled figure runs from 0.12 to 0.95 kg/m², which is wide relative to the effect itself. A number this modest, with a range this wide, is not a basis for expecting a specific outcome on your own body.

The trial title itself is a clue worth reading carefully: it evaluated metformin “with or without lifestyle modification” against placebo. That means the 0.53 kg/m² figure is not metformin measured against doing nothing — in many of the pooled trials, both arms had some level of dietary or activity counselling, and the number above reflects what metformin added on top of that, not what it does from a sedentary, unchanged baseline. The guideline’s own conclusion frames metformin as an efficacious adjunct to lifestyle interventions, not a substitute for them — the two are not separable in most of the evidence behind this number.

Is there a special metformin dose for PCOS weight loss?

No. Every trial behind the figures above used the same metformin doses prescribed for insulin resistance generally — there is no separate, higher, or different “weight-loss protocol” that the research supports. If you have seen a specific milligram number attached to weight loss specifically, it did not come from a trial that isolated dose as the reason for a bigger result on the scale. Metformin for PCOS covers what actually goes into that dosing decision — kidney function, why you are taking it, and gut tolerance — because the decision is the same one regardless of which outcome you are hoping metformin will move.

Why the effect on weight is this small

Metformin does not suppress appetite or increase energy expenditure — mechanisms that anti-obesity drugs are built around. It lowers the glucose your liver releases and improves how your muscle and fat tissue respond to the insulin you already produce. Weight change, where it happens, is a downstream effect of improved insulin sensitivity, not a direct action on hunger or intake. That is the mechanistic reason the number in Table 1 is measured in a kilogram or two, not several.

Side effects cut the effect short for a meaningful number of people

Gastrointestinal side effects are common enough to interrupt treatment before any metabolic benefit, weight-related or otherwise, has time to appear. A Cochrane review of 41 trials in 4,552 women found gastrointestinal side effects occurred in 22% to 40% of women on metformin, versus about 10% on placebo — an odds ratio of roughly 4. Nausea, cramping, and diarrhoea are usually worst in the first weeks and improve with a slow dose increase, food, or an extended-release formulation, but a real share of people stop before reaching that point. If the number on the scale has not moved and you also never made it through the early weeks of side effects, the trial average above was never given a fair chance to apply to you.

How metformin compares with other drugs tested for weight in PCOS

A 2018 network meta-analysis compared the drugs most often used off-label or on-label for weight in PCOS with obesity or overweight. Across 23 trials and 941 women, the amount of weight lost differed significantly among the drugs — in descending order, liraglutide, then orlistat, then metformin. Metformin came last of the three. Liraglutide and combined liraglutide-metformin also reduced waist circumference; metformin alone did too, but orlistat did not.

Newer GLP-1 drugs — semaglutide and tirzepatide, which postdate this 2018 analysis — produce substantially larger average losses in general-population trials than liraglutide did. GLP-1 drugs for PCOS covers what that class actually does and what happens when someone stops it, which is a separate and more consequential question than anything metformin raises. A third pharmacological option outside this comparison entirely — phentermine, an older stimulant-based appetite suppressant limited to short-term use — works through neither the GLP-1 nor the insulin-sensitising pathway.

Who this will not work for

If your primary goal is significant weight loss, metformin is not the tool the evidence supports reaching for first. The 0.53 kg/m² average in Table 1 means most people taking it will not notice a meaningful change on a scale, even though insulin resistance and lipid markers may still improve underneath that number. People expecting metformin to function like a GLP-1 drug or a structured weight-loss programme are working from the wrong evidence base, and the mismatch between expectation and reality is one of the most common reasons people stop taking it early, before the metabolic benefits it is actually good at have had time to show up.

Metformin’s weight effect is also not a measure of anyone’s effort or discipline, and it should never be read as one. PCOS itself drives weight gain through mechanisms — insulin resistance, altered appetite signalling, and androgen effects on fat distribution — that operate independently of how consistently someone eats or moves. Why PCOS causes weight gain sets out those mechanisms in full, and metformin only touches one of several of them, and only modestly.

A lean-PCOS reader is a distinct case worth naming directly. Trials behind the Table 1 average overwhelmingly enrolled participants with overweight or obesity, because that is where a metabolic drug has the most to act on. If your BMI is already in a typical range, there is little published data on what metformin does to your weight specifically, and the honest expectation is closer to “negligible” than “1 to 1.5 kg,” since the mechanism has less excess insulin resistance to correct.

You may see PCOS referred to as polyendocrine metabolic ovarian syndrome (PMOS) after a 2026 global consensus of more than 50 organisations renamed it. The metformin trials above were all conducted and reported under the older name; nothing about the drug or the numbers changed with the rename.

Common questions

If insulin resistance, not the scale, is what you actually want to address, metformin for PCOS covers the full trial picture, and the PCOS supplements guide ranks where non-prescription options in the rest of the supplements and medications section fit alongside it.